Schaffer Online Library of Drug Policy Sign the Resolution
Contents | Feedback | Search
DRCNet Home
| Join DRCNet
DRCNet Library | Schaffer Library | Drugs and Driving 


Carbohydrate-Deficient Transferrin (CDT) during 2 to 3 weeks of Alcohol Misuse in Men and Women and its Significance in View of Cut-Off Levels

T Gilg, J Deinl, H Grundner, L vMeyer, W Eisenmenger

Institute of Forensic Medicine, LM-University of Munich, Frauenlobstr. 7a, 80337 Munich, Germany


CDT has been described as a more sensitive and especially specific marker of recent heavy alcohol consumption than others like GGT. Whereas specifities of at least 90% are reported, sensitivities are poorer ranging from 22 to 79%. Only a few investigations refer to gender differences with lower specifities in females. The course of CDT (incl. methanol, isopropanol and GGT) was investigated during 2 resp. 3 weeks with a daily ethanol consumption of 50 to 60g in less than 3 hours in the evenings in - so far - 7 female and 12 male healthy persons. CDT was determined in duplicate by a double antibody kit (CDTectR, Kabi Pharmacia).

Methanol and isopropanol as shorttime as well as GGT as medium-term marker did not react. CDT increased in 8 of 12 males, with 5 exceeding the cut-off level of 20U/l and 3 cases supposed to exceed cut-off levels after a prolonged alcohol consumption of more than 3 weeks. 4 showed no reaction (non responders). In contrast to recent studies, females also showed an increase in 6 of the 7 cases, with 3 exceeding the upper normal level of 26U/l after 10 days and 3 cases supposed to surpass it after prolongation to 14 days. One female did not respond.

According to these preliminary results CDT provides an important instrument for detection of alcohol drinking of at least 60g/day over a minimum of 3 weeks also in unselected male and female populations for use in clinical and traffic medicine.


Detecting heavy drinking, alcohol abuse or alcoholism resp. alcohol dependance is of considerable interest not only in clinical, but also in traffic medicine. Questionnaires such as CAGE or MALT require cooperation, which may not be given in subjects regranting licences after questioned general driving ability related with alcohol problems. Biological markers are useful to detect excessive or chronic alcohol consumption, although the diagnosis of alcoholism solely by markers is limited. Trait markers (e.g. monoaminoxidase-B-activity, adenylatcyclase, endocrine markers, dopamin-beta-hydroxylase, evoked potential P 300 and ADH/ALDH-Phenotypes) represent a genetic predisposition to alcoholism, but have not been firmly established. State markers as mainly metabolic effects of alcohol are widely used for diagnosis and screening, e.g. GGT, ASAT, ALAT, HDL-Cholesterin, MCV and markers for recent heavy alcohol consumption like b-hexosaminidase, urinary 5-HTPOL/5-HIAA, isopropanol/acetone and especially methanol (overviews in Crow/Batt 1989, Watson 1989, Taberner/Badawy 1992, Iffland et al 1994, Seitz/Lieber/ Simanowski 1995, Gilg 1995, Gilg/Soyka 1995, Kranzler 1995). Both specifity (% of negative tests in individuals without the disease resp. false positives) and sensitivity (% of abusers with positive or abnormal tests resp. false negatives) are limited in most known markers.

According to many researchers, carbohydrate-deficient transferrin (CDT) - an abnormal transferrin characterized by lowered concentrations of mainly sialic acid resp. N-acetyl-neuraminic acid, resulting in tri-, di-, mono- or asialotransferrins - is a fairly sensitive, but outstandingly specific marker for excessive alcohol consumption. Several studies point at CDT as the most specific marker of alcoholism or - reduced to the actual meaning - of an average daily consumption of at least 60g ethanol, with specifities up to 99% resp. always exceeding 90% (Stibler 1991 and Allen et al 1994, Arndt et al 1994). CDT is well suited for a discrimination between alcohol-induced hepatopathy and liver diseases of other origin (Xin et al 1992, Bell et al 1993, 1994, Yamauchi et al 1993) and for detection of relapse (Carlsson et al 1993). Only a few serious nonalcohol-related liver diseases like primary biliary cirrhosis, chronic active hepatitis, severly decompensated liver cirrhosis and advanced cirrhosis with ascites may elevate CDT, as well as rare cases of atypical transferrin variants and an inherited disorder - carbohydrate deficient glycoprotein syndrome or CDG (Stiebler 1991, Heyne and Weidinger 1992, Allen et al 1994).

The predictive values for alcoholism derived from CDT are particularly high (with sensitivities up to 94%), when applied to the identification of male alcoholics (Behrens et al 1988, Gjerde et al 1988, Stibler 1991). However, the efficacy of CDT to detect alcohol abuse or problematic drinking in other populations like collage-aged students, particularly females, in general medical or community setttings as well as in female alcohol abusers seems to be limited, with sensitivities as low as 22% (Nilssen et al 1992, Nyström et al 1992, Anton and Moak 1994, Anton and Bean 1994, La Grange et al 1994, 1995, Allen et al 1994). Unless many data indicate, that a consumption of >50-80g of ethanol/day for at least one week is required to produce this abnormality in between 81% and 94% of patients with alcohol-related symptoms (Stibler 1991, Allen et al 1994), there are only a few reports on the amount of alcohol required to produce elevated CDT levels, especially with respect to the rise of CDT in controlled drinking over 2 to 3 weeks and in normal healthy and especially female subjects. Salmela et al (1994) found an increase of CDT in 10 male volunteers after 60g ethanol/day for 3 weeks, but only 2 exceeded the upper normal level of 20 units/liter (U/l). For further evaluation we studied the concentration-time course of CDT, GGT, methanol and isopropanol/acetone in 12 male and 7 female subjects during 10 to 21 days of alcohol consumption.


After abstaining from alcohol for one week, 7 female and 12 male volunteers (healthy staff members and their relatives, moderate drinkers without any alcohol history) ingested beer or wine within 3 hours at late afternoon or early evening in doses equivalent to 60g ethanol/day for 10 up to 21 days (details in Table 1). Blood samples were collected in the mornings at the beginning and thereafter every two or three days. Serum was separated and analyzed for CDT as well as GGT and methanol (incl. ethanol, isopropanol, acetone and other congeners). Methanol and congeners were analyzed according to our specialized, routine head-space gaschromatography. CDT was determined in duplicate by a double antibody RIA kit (CDTectR, Kabi Pharmacia AB Diagnostics, Uppsala, Sweden) in our laboratory according to the manufacturer´s instruction. Coefficients of variation in high and low levels, within day and from day to day were not exceeding 10%. Values were stable on storage of serum at -20°C for at least some months, at -4°C for up to 3-4 weeks. Full blood samples were stable only a few days and only in upright position to prevent major contamination of serum with hemoglobin. (Autoptic blood samples showed regular results with postmortal intervals up to two days).

Table 1
Subjects and Conditions of Ethanol Consumption

  age (years)
mean (range)
weight (kg)
mean (range)
ethanol load
ethanol load
ethanol load
(g/kg x r/d)
female (n=7) 37 y
62 kg
50 g
0,86 g
1,43 g
male (n=12) 31 y
78 kg
60 g 0,78 g
1,11 g


As expected, isopropanol/acetone and esp. methanol as a marker for recent heavy alcohol consumption resp. longtime alcoholization of 1-2 days did not react and showed typical endogenous levels in the range of 1mg/l (Iffland et al 1994, Gilg 1995).

GGT as a marker for ethanol consumption over weeks also did not react, not even with a rise in between the normal range. One male and one female subject initially showed elevated GGT-levels exceding the cut-off, followed by a more or less even course. In both cases there was absolutely no history or other signs of alcohol abuse, the female volunteer was even non responding with CDT during the alcohol period.

In 8 of the 12 male subjects, CDT showed an increase over the drinking period, with 5 of them exceeding the cut-off, whereas in 4 subjects a distinct increase was not given. One subject underwent a second phase of ethanol load after some weeks, then with a daily consumption of alcohol as a bolus within one hour and for 14 days. With a drinking period of 3 hours, the blood alcohol concentration (BAC) did not exceed 0.5o/oo, whereas within one hour a BAC over 0.5 up to 0.8o/oo can be reached, which is better suited to inhibit other enzymatic processes. There are no differences, therefore it does not seem, that BAC of more than 0.8o/oo have to be reached to elevate CDT. Also the course of CDT indicates, that a prolongation of ethanol intake could resp. should lead to levels exceeding the cut-off after at least 3 weeks of chronic alcohol consumption.

Two subjects were observed, each with 90 kg body weight and exactly the same pattern and kind of alcohol consumption. Whereas the one with an age of 24 years showed a nearly continuous rise, the 45 year old did not react, not even with an elevation within the normal range.

Seven female subjects, with similar reactions and cut-offs reached after prolongation closer to 2 than to 3 weeks.

Table 2 summarizes the results. 3 of 7 resp. 43% of the female subjects exceeded the upper normal level of 26U/l already after 10 days of drinking, whereas the male subjects mostly needed 2 to 3 weeks. One reason could be the higher systemic and hepatic ethanol load, because of the lower volume of distribution in females resulting in higher BAC (Tables 1, 2). Extrapolating the courses of CDT under the presumption of a more prolonged drinking would give an exceed of cut-offs in 86% in females after 2 weeks resp. 67% in males after 3 weeks.

Table 2
CDT after a Daily Consumption of 50/60g Ethanol for 2/3 weeks

  ethanol load (g/d) ethanol load (g/kg x r/d) duration (days) non- responder cases > cut off (>26/20U/I) cases > cut off > 2/3 weeks cases > cut off total
female (n=7) 50 g (60) 1,43 g 10 1 3 = 43% 3 = 43% 6 = 86%
male (n=12) 60 g 1,11 g 14 (10/21) 4 5 = 42% 3 = 25% 8 = 67%


Our results are similar to those from the only comparable study from Salmela et al (1994), who found an increase of CDT in 10 male volunteers after 60g/day for 3 weeks. In their study, only 2 of the 10 subjects CDT levels exceeded the cut-off as defined by the manufacturer, compared to 5 out of 12 in our investigation. One of their curves gives the impression of a non responder, whereas we found 4 such cases. Considering, that Salmela et al also used CDTectR, differences in individuals or drinking patterns are possible reasons. E.g. they had weekly intervals of blood taking and did not mention the drinking pattern, a consumption spread over the day may render different results because of lower BAC and reduced metabolic effects, also women were not investigated. The quantitative CDTectR, isoelectric focusing methods with differentiation of isoforms as well as the relative Axis % CDT test from Bio Rad generally show good correlations, moderate differences in measurements and accuracy may only partially explain different outcomes especially in women (Lof et al 1993, Sillannaukee et al 1993, Bell et al 1994, Allen et al 1994). Some researchers as Nilssen et al (1992) and Anton and Bean (1994) reported, that CDT performed more poorly as a marker of heavy alcohol consumption in women than in men, e.g. with about half the values for sensitivities, but irrespective of assay procedure. Our first longitudinal survey of CDT in women with defined ethanol load indicates a sensitivity comparable or even better than in men. Possible explanations may be different subject characteristics, age, drinking patterns or body weight and volume of distribution, which should be elucidated by further research resp. higher numbers. Reasons for increased levels of CDT in normal, socially drinking women with cut-off levels of 26U/l are not clear. Results on influences by pregnancy, hormonal issues, oral contraceptives or total serum iron content are controversely (Allen et al 1994, La Grange et al 1995). The recommended cut-off levels of 20U/l (partly 22 or 28U/l) in men and 26U/l (-27U/l) in women are based on subjects without or with moderate consumption of ethanol (15/20-40g/day). According to Iffland et al (1994), Kanitz et al (1993) and Arndt et al (1994) a general cut-off level of 30U/l (as a 3s-value) could be safe even for forensic purposes. In view of our results, the recommended levels of 20 and 26U/l should be sufficient, if an exact analysis is provided (which may be even better with new enzyme immuno assays, announced for April 1995). Based on the high specifity, this proves a misuse of alcohol for at least 2 to 3 weeks before the test, but otherwise cannot exclude even 60g/day in some cases. For screening purposes, the relatively fast biological turnover has to be considered. The few cases, where a distinctive decline resp. turnover was obvious, it was in the range as described for alcoholics after withdrawal, that means half-life times of normalization of 14 -17 days (Stibler 1991, Kanitz et al 1993, Allen et al 1994) resp. 12.0 ± 3.4 days (n = 36, own results, Gilg 1995).

Methanol as a marker for recent alcohol intake (at least hours up to 1-2 days of periods of alcoholization with > 0,2-0,5o/oo) and GGT (needs some weeks of alcohol intake) did not react, which symbolizes the different patterns of alcohol consumption and corresponds with reports that CDT may coincide, but does not correlate with other markers (Anton and Moak 1994, Iffland et al 1994, Gilg et al 1994, Salmela et al 1994).


Our results show, that CDT detects chronic daily alcohol intake equivalent to 60g ethanol in non-alcoholic subjects in 3 out of 7 women (after 10 days) and 5 of 12 men (after 14/21 days). Furthermore they indicate, that CDT may detect 6 out of 7 women after 2 weeks and 8 of 12 men after 3 weeks, if consumption of alcohol is continued. Especially in women the sensitivity to detect an alcohol abuse is higher than previously described. Due to the known high specifity, CDT levels exceeding cut-offs strongly point at alcohol abuse eqivalent to at least 60g/day over the preceding 2 resp. 3 weeks. Otherwise there are non responders, so that even normal CDT-values cannot exclude alcohol abuse in this manner. Further controlled trials should be conducted among non-alcohol-dependent individuals to better relate quantity and pattern of drinking to subsequent CDT levels, esp. minimal level or duration of drinking, and to evaluate influences of gender, age, body weight and reasons for non responding.


Allen JP, Litten RZ, Anton RF, Cross GM: Carbohydrate-deficient transferrin as a measure of immoderate drinking: Remaining issues. Alcohol Clin Exp Res 18,799-812 (1994).

Anton RF, Moak DH: Carbohydrate-deficient transferrin and -glutamyltransferase as markers of heavy alcohol consumption. Alcohol Clin Exp Res 18/3,747-754 (1994).

Anton R, Bean P: Two methods for measuring carbohydrate-deficient transferrin in inpatient alcoholics and healthy controls compared. Clin Chem 40/3,364-368 (1994).

Arndt T, Gressner AM, Kropf J: Labordiagnostik und Kontrolle des Alkoholabusus. Med Welt 45,247-257 (1994).

Behrens U, Worner TM, Braly LF et al: Carbohydrate-deficient Transferrin, a Marker for Chronic Alcohol Consumption in Different Ethnic Populations. Alcohol Clin Exp Res 12,427-432 (1988).

Behrens UJ, Worner TM, Lieber CS: Changes in Carbohydrate-Deficient Transferrin levels after alcohol withdrawal. Alcohol Clin Exp Res 12,539-542 (1988).

Bell H, Tallaksen C, Sjahem T, Weberg R et al: Serum carbohydrate-deficient transferrin as a marker of alcohol consumption in patients with chronic liver diseases. Alcohol Clin Exp Res 17/2,246-252 (1993).

Bell H, Tallaksen CME, Try K, Haug E: Carbohydrate-deficient transferrin and other markers of high alcohol consumption: A study of 502 patients admitted consecutively to a medical department. Alcohol Clin Exp Res 18,1103-1108 (1994).

Bell H, Tallaksen CCM, Haug E, Try K: A comparison between two commercial methods for determining carbohydrate deficient transferrin (CDT). Scand J Clin Lab Invest 54, 453-457 (1994)

Carlsson AV, Hiltunen AJ, Beck O, Stibler H, Borg S: Detection of relapses in alcohol-dependent patients: comparison of carbohydrate-deficient transferrin in serum, 5-hydroxytryptophol in urine, and self-reports. Alcohol Clin Exp Res 17,703-708 (1993).

Crow KE, Batt RD (ed): Human metabolism of alcohol. Vol I: Pharmacokinetics, medicolegal aspects and general interest. Vol II: Regulation, enzymology and metabolites of ethanol. Vol III: Metabolic and physiological effects of alcohol. CRC Press, Boca Raton, Florida, (1989).

Gilg T, Weidinger S, Josephi E, Tutsch-Bauer E, vMeyer L, Bieger WP: Bestimmung von CDT (Carbohydrate Deficient Transferrin), -GT, Methanol, Isopropanol und Ethanol in forensischen Blutproben zur Beurteilung chronischen Alkoholmißbrauchs. Lab Med 18, 143-153 (1994)

Gilg T: Stellenwert von Begleitstoffen (Methanol, Isopropanol) und CD-Transferrin in der Alkoholismusdiagnostik. Kap.6 in: Biologische Alkoholismusmarker (Roundtable 23.7.94 in Munich, ed.: M. Soyka), Chapman and Hall, Weinheim (1995)

Gilg T: Diagnose von Alkoholmißbrauch und Alkoholismus, biologische und biochemische Alkoholismusmarker bzw. Parameter. In: Die Alkoholkrankheit - Diagnostik und Therapie (ed.: M. Soyka), Chapman and Hall, Weinheim, (1995).

Gjerde H, Johnsen J, Bjoerneboe GE, Moerland J: A comparison of serum carbohydrate-deficient transferrin with other biological markers of excessive drinking. Scand J Clin Lab Invest 48,1-6 (1988).

Heyne K, Weidinger S: Diagnostik und Nosologie der Glykanose CDG ("Carbohydrate-deficient glycoproteinsyndrome. Monatsschr Kinderheilk 140, 822-827 (1992)

Iffland R, Balling P, Börsch G et al: Zur Wertung erhöhter Spiegel von GGT, CDT, Methanol, Aceton und Isopropanol im Blut alkoholauffälliger Kraftfahrer. Blutalkohol 31,273-314 (1994).

Kanitz RD, Wetterling T, Missler U: Carbohydrate deficient transferrin als Indikator zur Objektivierung eines pathologisch erhöhten Alkoholkonsums. Fortschr Diagn 4 (Praxis Report 5), (1993).

Kranzler HR (editor): The pharmacology of alcohol abuse. Springer-Verlag, Berlin-Heidelberg New York, (1995).

LaGrange L, Anton RF, Crow H, Garcia S: A correlational study of carbohydrate-deficient transferrin values and alcohol consumption among hispanic college students. Alcohol Clin Exp Res 18/3,653-656 (1994).

LaGrange L, Anton RF, Garcia S, Herrbold C: Carbohydrate-deficient transferrin levels in a female population. Alcohol Clin Exp Res 19/1,100-103 (1995).

Lof K, Koivula T, Seppa K, Fukunaga T, Sillanaukee P: Semi-automatic method for determination of different isoforms of carbohydrate-deficient transferrin. Clin Chim Acta 217, 175-186 (1993)

Nilssen O, Huseby NE, Hoyer G et al: New alcohol markers-How useful are they in population studies: The Svalbard study 1988-89. Alcohol Clin Exp Res 16,82-86 (1992).

Nyström M, Persalo J, Salaspuro M: Carbohydrate-deficient transferrin (CDT) in serum as a possible indicator of heavy drinking in young university students. Alcohol Clin Exp Res 16,93-07 (1992).

Salmela KS, Laitinen K, Nyström M, Salaspuro M: Carbohydrate-deficient transferrin during 3 weeks' heavy alcohol consumption. Alcohol Clin Exper Res 18/2,228-230 (1994).

Seitz KH, Lieber CS, Simanowski UA (ed.): Handbuch Alkohol Alkoholismus Alkoholbedingte Organschäden. Barth, Leipzig Heidelberg, (1995).

Sillanaukee P, Seppa K, Lof K, Koivula T: CDT by anion-exchange chromatography followed by RIA as a marker of heavy drinking among men. Alcohol Clin Exp Res 17,230-233 (1993).

Stibler H: Carbohydrate-deficient transferrin in serum: a new marker of potentially harmful alcohol consumption reviewed. Clin Chem 37,2029-2037 (1991).

Watson RR (ed.): Diagnosis of alcohol abuse. CRC Press, Boca Raton, Florida (1989)

Xin Y, Rosman AS, Lasker JM, Lieber CS: Measurement of carbohydrate-deficient transferrin by isoelectric focusing/western blotting and by micro anion-exchange chromatography/radioimmunoassay: comparison of diagnostic accuracy. Alcohol Alcohol 27,425-433 (1992).

Yamauchi M, Hirakawa J, Maezawa Y et al: Serum level of carbohydrate-deficient transferrin as a marker of alcoholic liver disease. Alcohol Alcohol Suppl 1B,3-8 (1993).